Dysfunctional Root Systems and Brief
Landscape Lives: Stem Girdling Roots and
the Browning of Our Landscapes
Gary Johnson
Consider this comparison of potential life spans for trees (Burns and Honkola 1990; USDA 1998)
Quercus macrocarpa (bur oak), in upland site
250+ years
Acer saccharinum (silver maple), in riparian site
125+ years
Acer negundo (boxelder), in lowland site
100+ years
Pinus banksiana (Jack pine), in field site
80+ years
Betula papyrifera (paper birch), in northern lowland forest
65+ years
Tree planted in urban core street site
T
hat’s a sobering thought—a tree with a
normal life span of 65 to 250 years may
live less than 10 years when planted in any
American city’s downtown landscape. Admittedly, that figure represents tree placement in
the worst of our urban landscape sites: sidewalk
cut-outs. These inhospitable planting sites are
also known as tree coffins, tree burial mounds,
or urban tree disposal units to frustrated urban
foresters. When the mortality rate of downtown
trees is compared to tree losses from Dutch elm
disease, oak wilt, sudden oak death, and gypsy
moth, it doesn’t take too long to realize that
there’s an epidemic of urban tree loss going on
and it’s largely under the radar (Figure 1).
Another oft-quoted number is that the average urban residential tree lives for 30 to 35 years
(Moll 1989). That life span is three times as
long as a sidewalk tree, yet only half as long as
a paper birch in its natural environment. Growing conditions in residential landscapes may
not be quite as bad as sidewalk sites, but there
are many natural and unnatural pressures on
the trees that lead to briefer landscape lives.
Residential landscape soils can be as stressful as
downtown sites: poorly drained, outrageously
alkaline, subjected to blends of every pesticide
less than 10 years
known to modern society, and compacted to
such a degree that lawns may seem like nothing
more than green concrete.
With few exceptions (perhaps tornadoes and a
few diseases), there are no “angels of death” that
descend and quickly kill trees in landscapes.
More commonly, a multitude of predisposing
stresses that occur in our highly altered urban
landscapes combine to weaken trees over the
years. Often, inciting events such as floods or
hailstorms and/or contributing agents such as
target cankers or wood boring insects complete
the job for the majority of tree losses. Meanwhile, plant health care professionals attempt
to determine the true causes of decline and
death, and often the diagnoses are incomplete
or incorrect because of the multiple offenders
involved with the problem.
Predisposing Factors and Tree Decline
When trees are chronically stressed (long-term
drought, repeated defoliation, etc.), their normal reserves of chemical energy—primarily as
complex carbohydrates—are slowly depleted.
Each year as stressed trees come out of dormancy, they emerge in a weakened state due
to this energy depletion and find it increas-
Photo by Gary Johnson
Dysfunctional Root Systems 3
Figure 1. Trees in urban sidewalk sites are subjected to very unhealthy environments and live less than 10 years on average.
ingly difficult to releaf, grow, and deal with the
harsh realities of urban landscapes on a normal
basis. It takes a tremendous amount of chemical energy to push out new leaves and shoots,
recover from accidental wounds on the stems,
or produce flowers and fruit.
As the tree’s energy reserves continue to
decline—and thereby affect the tree’s ability
to capture and store new energy through photosynthesis—the entire system is affected and
the decline spiral to premature death begins.
So decline in a sense refers to the tree’s ability
to deal with life’s normal stresses. A tree in
decline may die suddenly because of an event
such as a cold winter with no snow cover, a
short-term summer drought, or a defoliation
from insects or hail. The other trees in the landscape tolerate the damage and survive, but the
predisposed trees—those in decline—are unable
to recover from the damage.
Dysfunctional Root Systems as
Predisposing Agents
Despite the fact that roots are seldom seen, dysfunctional root systems are too often the predisposing agents connected to tree health decline,
and ultimately the reason why many urban
landscape trees experience such brief lives. If the
root system—approximately 50% of a tree’s biomass—is not operating normally, the entire system will be abnormal. Abnormal is not always
harmful, as seen in bonsai plants and trees growing on slopes. In bonsai plants, a restricted root
system causes compacted growth in the rest of
the plant system, but the system itself may be
healthy and completely functional under most
circumstances. In the case of a tree growing on
a slope, the tree is anchored with a skewed and
asymmetrical root system, but its overall health
is not compromised even though the root system
could certainly be considered abnormal.
which create a root system so dysfunctional that it can end up killing
the entire tree.
Stem Girdling Roots as
Predisposing Agents
Stem girdling roots are those roots
that grow either partially or completely against the tree’s stem and
compress (girdle) the stem tissues
(Figure 2). Xylem and phloem tissues
in the stem become much narrower
at the point of compression, impeding normal water movement and
sap flow (Figure 3). This restriction
affects energy reserves by directly
Figure 2. With part of a stem girdling root removed, the compression to the
and indirectly affecting photosyntree’s trunk is evident.
thesis. Trees become stressed and
But abnormal root systems that do affect
more vulnerable to secondary problems. For
the overall health or stability of the tree are
this reason, SGRs are considered to be primary
considered dysfunctional. For example, when a
predisposing agents in landscape tree decline
container-grown tree with a severely pot-bound
and death.
root system is planted, its rhizosphere does not
Some of the first symptoms of SGR-impacted
occupy a large enough area to capture sufficient
tree health include leaf scorch or leaf wilting
water and nutrients needed to support a normal
on a tree when no other plants in the area are
sized tree without supplemental help. Dysfuncshowing the same symptoms. There may be
tional root systems are also common on newly
adequate moisture in the soil, but the tree’s
transplanted bare-root and balled-and-burlapped
ability to move water throughout the system is
plants; these plants often lose 75% or more of
thwarted by the areas of compression, i.e. the
their root systems during the harvest operation,
greatly reduced diameter of vessel elements.
resulting in transplant shock which may go on
Soon, this water stress evolves into early leaf
for several years until the root system regrows.
coloration and leaf drop in the summer, late
And then there are stem girdling roots (SGRs),
leaf-out in the spring, and chlorosis or other
G. W. Hudler, Cornell University
Photo by Dave Hanson
4 Arnoldia 66/2
Figure 3. Transverse views of normal Norway maple stem wood showing a healthy growth pattern (left), and malformed stem wood
compressed by a stem girdling root (right). Water and nutrient transport in trees is negatively affected when tissue is malformed by
compression. V = vessel element, R = ray, F = fiber tracheid. Both views are at the same scale.
nutrient deficiency symptoms. If
the stem compression becomes
more severe, affecting 50% or
more of the stem circumference,
so do the symptoms. Trees will
tend to suffer more damage during
the winter seasons, in particular
true frost cracks, cambial death,
and dieback. In the latter stages of
decline due to SGRs, trees usually
suffer from severe stunting (very
small leaves, annual twig growth
of 1 to 2 inches or less) and significant defensive dieback. With Figure 4. The middle littleleaf linden was in the last stages of decline from stem
so little vascular capacity left, girdling roots at the time of this photograph. One year later it was dead.
affected trees may succumb completely from even a short-term
summer drought (Figure 4).
Though often a slow-acting
cause of death, SGRs can also
cause tree death that is a bit more
sudden and dramatic. The compressed areas of tree stems are
structurally weak points and far
too often are the points of failure during windstorms (Figure
5). For example, in severe windstorms that occurred in Minnesota in 1998, 73% of the lindens
(Tilia spp.) that were lost in urban Figure 5. Stem compression from SGRs located 4 or more inches below ground
landscapes failed at compression was the most common cause of urban tree failure in windstorms in Minnesota
points from SGRs, and most broke from 1995 to 2005.
several inches below ground. This
these SGRs were well below ground (from 4 to
is a different type of predisposition but equally
14 inches)—out of sight, out of mind (Figure 6).
damaging to a tree’s ability to grow, survive, and
In landscape surveys conducted by the Uniadd to the quality of life.
versity of Minnesota Department of Forest
More (Soil) is Not Always Better
Resources (1997 to 2004), five species of trees
were investigated in three different communiEarly SGR studies conducted by the University of
ties. All trees were growing in public spaces:
Minnesota were in response to unexplained tree
boulevards, schools, government centers, parks.
decline in urban areas. From 1994 through 1996,
Species surveyed included hackberry (Celtis
220 declining and dying trees were diagnosed. In
occidentalis), littleleaf linden (Tilia cordata),
81% of the cases, stem girdling roots were the
sugar maple (Acer saccharum), ‘Shademaster’
only causal agents isolated. This figure closely
honey locust (Gleditsia triacanthos ‘Shademasparalleled data collected from a national survey of
ter’), and green ash (Fraxinus pennsylvanica).
tree care professionals (Johnson and Hauer 2000).
Trees were randomly selected, evaluated for
More specifically, these trees had been planted
health and condition, and then examined for
in the previous 12 to 20 years and had signifidepth of soil over the main order roots and the
cant stem compression (greater than 50% of the
presence of stem encircling roots (potentially
stem circumference) from SGRs. In all cases,
Photos by Gary Johnson
Dysfunctional Root Systems 5
6 Arnoldia 66/2
Photos by Gary Johnson
conflicting roots within 6 inches of the stem)
or stem girdling roots. The results were a bit
depressing. Only 4% of the lindens, 8% of the
ash, 10% of the maples, 15% of the honey
locust, and 40% of the hackberries had their
stems completely above ground. The rest of
the sampled trees had from 1 to 12 inches
of soil over the first main order roots and
against the stems.
Non-destructive root collar examinations
were performed on a total of 1,380 trees. The
intent of these examinations was to determine
the frequency of SERs (stem encircling roots—
those potentially conflicting roots within six
inches of the stem) and SGRs associated with
different depths of soil (up to 12 inches) over the
first main order roots. The excavations demonstrated that the deeper tree stems were buried in
the soil or mulch, the more likely
it was for them to have multiple
layers of stem encircling and
stem girdling roots. The increased
presence of these problem roots
showed up in trees beginning with
as little as one inch of excess soil
against the stem. In a nutshell,
the more soil or pre-soil (organic
mulches that will break down)
that is piled over the root systems
and against the stems, the more
likely it is that trees will decline
or fail due to multiple conflicts
with SGRs (Figure 7).
Figure 6. This SGR, located approximately 4 inches below ground, runs tangential to the tree trunk and is compressing 30% of the stem circumference.
Figure 7. As shown on this littleleaf linden, more layers of SGRs develop as the
stem is buried deeper. Greater than 40% of the stem circumference of this tree
was compressed by several layers of SGRs.
How SGRs Form
Observations from the 1,380 root
collar examinations conducted
during the species surveys and a
separate nine-year planting depth
study have led to the conclusion
that stem girdling roots form in
one of two ways: first, new roots
regenerating from deeply buried
main order roots, and second,
from stem adventitious roots.
When main order roots are buried
too deeply, new woody roots that
originate from them or any part
of the buried root system tend to
grow closer to the surface. It is
speculated that this action is in
response to a more desirable soil
oxygen and moisture balance. As
the roots reach the soil surface,
an unpredictable percentage of
them grow tangential to the tree
stem or in some cases encircle
the stem. For the next number of
years (12 to 20, from our observa-
Dysfunctional Root Systems 7
tions), the roots and stems expand in diameter,
resulting in the ultimate confrontation between
roots and stems.
Stem adventitious roots are also sources
of SGRs. When a buried stem begins forming
adventitious roots, many or most of those roots
grow away from the stem in a radial fashion.
As with new roots growing from main order
roots, an unpredictable percentage of these
adventitious roots do not grow radially but
instead grow tangential to the stem or encircling the stem. The interface area between soil
and stem appears to be a highly desirable area
for stem root growth, perhaps because it provides an ideal balance of soil oxygen and moisture and is also the path of least resistance for
root proliferation. The exact reasons for these
root growth responses are still speculative, but
it is clear that when tree stems are buried by
a media that supports root growth, SGRs are
highly likely to occur.
It’s worth noting that stem girdling roots are
a problem primarily with younger trees. As
trees mature, their growth slows down dramatically, including the growth of trunk diameter
and encircling roots. Because of this reduced
growth—and the fact that there is often a relatively thick outer bark—stems of mature trees
that then become buried by soil or organic matter are much less likely to develop stem girdling root problems. SGRs can still develop, but
if they do they are less likely to result in the
decline and death of the tree.
How to Cause Stem Girdling Roots
If you want to cause the formation of SGRs,
bury the tree stem with a medium that supports root growth. Here are some common ways
SGRs occur:
•Excess soil is piled over the first main order
roots during the growing and harvesting of
balled-and-burlapped trees.
•Excess growing medium buries stems when
container-grown trees are up-potted.
•Decayable organic mulch is piled high
around tree stems in nurseries and landscape sites.
•Soil is piled against tree stems during construction regrading in landscapes.
•Trees are planted in a new landscape before
final grading is completed.
There are so many different ways that stems
can be buried—accidentally or with good
intentions—that it is difficult to pinpoint
the main source of the problem. One seemingly common cause is the act of burying trees
rather than planting trees. Unfortunately, too
many people still have the notion that trees
are like fenceposts and need to be buried deep
for stability. Not so.
In 2002, we conducted a planting depth
study in collaboration with a large wholesale nursery. Bare-root birch (Betula spp.), ash
(Fraxinus spp.), and crabapple (Malus) were
potted up in number-ten containers at four
different depths: 0, 2, 4, or 6 inches of soil
over the first main order roots. On a weekly
basis, each of the 240 trees was inspected for
lean or windthrow from the containers. At the
end of the four month study, all trees were
well-rooted in the containers and the results
of the study showed that all trees, regardless
of depth, leaned at the same frequency and to
the same degree. Planting tree stems deeper
had absolutely no positive effect on tree stability. If newly planted trees are unstable, they
may need temporary support from a guying or
staking system, not entombment.
Nine Years of Burial
In 2000, a long-term planting depth study was
installed at the University of Minnesota’s Urban
Forestry and Horticulture Institute’s research
fields. Three hundred and sixty trees equally
represented by two species (sugar maple [Acer
saccharum] and littleleaf linden [Tilia cordata])
were planted at three depths: 0, 5, or 10 inches
of soil over the first main order roots. All trees
were planted in a complete, randomized block
design in a .75 acre plot as unbranched, 2 to
3 feet tall liners. At three year intervals, onethird of the trees were harvested and had their
root systems excavated with a supersonic air
tool. Each year, mortality rates, growth rates
(stem caliper), number of suckers produced, and
percentage of dieback was recorded. In 2009,
the final third of the original experiment will
be harvested, but some interesting trends and
Photo by Nancy Rose
Photo by Gary Johnson
8 Arnoldia 66/2
Figure 8. Bury the stem of littleleaf linden just 5 inches deep and a profusion of suckers will develop. These suckers eventually
become SGSs (stem girdling suckers) as they grow in caliper and compress the tree’s stem.
significant data have already been revealed from
the first two harvests, including:
•Planting sugar maples 5 to 10 inches too
deep is an effective way to kill them. The
mortality rates for the 0, 5, and 10 inch
depths as of 2006 were 30, 40, and 65%,
respectively.
•There was a significant positive relationship between placing 5 to 10 inches of soil
against the stems and the frequency of SGRs
on Tilia cordata in both the 2003 and 2006
harvests. Acer saccharum showed a trend in
the same direction.
•Tilia cordata with stems buried in 5 inches
of soil will produce masses of stem suckers, making the tree look more like a shrub.
Sucker formation on Tilia cordata doesn’t
just ruin the tree’s appearance, it can also
cause premature failure. Stem girdling suckers (SGSs) are suckers that form prolifically
and, when they enlarge in diameter, can
girdle the stem vertically and horizontally
(Figure 8).
How Often do Trees Die from SGRs?
This question is likely unanswerable. When
trees suddenly fail and die during a windstorm,
diagnosing the problem below ground is not
often considered. Weather alone is often blamed
for the deaths, and the trees are hastily removed
and replaced.
Research we conducted from 1995 through
2005 on tree failure in windstorms exposed a
Dysfunctional Root Systems 9
the exact percentage of littleleaf linden that
failed during the previously mentioned 1998
storms. After 11 years of data collection, the
presence of SGRs and, more specifically, stem
compression from SGRs that amounted to 50%
or more of the stem circumference, emerged as
the number one reason why urban trees failed
in windstorms.
What to Do, What to Do?
Prevention is the easiest and most effective way
to eliminate the SGR problem in landscapes.
Whether you are an urban forester, commercial
landscaper, or home gardener, follow these steps
to prevent or manage stem girdling roots:
•Don’t plant container or balled-and-burlapped trees that are already buried too
deeply. Assume there is too much soil over
the first main order roots and remove that
excess soil before planting a newly purchased tree (Figure 9).
•Plant trees, don’t bury them. If stems aren’t
buried, it’s not likely that SGRs will become
a problem. They can still occur on correctly
planted trees, but much less frequently than
on buried trees.
Photo by Gary Johnson
broader picture of the effects SGRs have on
landscape trees. During this period over 1,500
“tree autopsies” were conducted on trees that
had failed during wind-loading events in Minnesota. These trees were not those from the
centers of severe wind-loading events such as
straight-line winds or tornados. Rather, they
were victims of thunderstorms or those at the
edges of severe wind events.
From that data, the destruction and economic
losses from premature tree failures due to SGRs
were determined, and it was startling. The most
common tree size category for boulevard tree
failures was the 6 to 10 inch DBH (diameter
at breast height, 4.5 feet above ground) range.
Of those trees, 50% snapped off at compression points from SGRs at a depth of 4 or more
inches below ground. The Achilles’ heel was
a compression root that couldn’t even be seen
because the stem was buried so deeply. The data
also indicated that littleleaf lindens (Tilia cordata) were grossly affected by SGRs. Littleleaf
linden ranked as the third most common species for total failure (the tree went down completely) during those years, and 73% of those
trees snapped off at below-ground SGRs, almost
Figure 9. Most containerized trees will have 2 to 6 inches of excess soil over the first main order roots and against the
stem. Use a pruning saw to remove this excess soil before planting. Of 500 trees subjected to this treatment at the
University of Minnesota’s research nursery, there has been a 0.7% mortality rate in 2.5 years.
•Don’t pile mulch
against stems. Organic
mulch is basically presoil. Piling on mulch
will result in a buried
stem and a wonderful
environment for SGRs
to develop.
•When suspicious,
investigate. Root collar
exams are not all that
difficult to perform
(Figure 10). If you have
a trowel and a wet-dry
vacuum, you can perform a non-destructive
root collar exam. If
you find offending
roots during the exam,
remove them. Also,
Figure 10. The fastest and most non-destructive method for conducting a root collar exam
remove all that extra
is with a supersonic air tool that blows the soil away without harming the roots. This root
soil. If you do nothing,
collar exam was accomplished in approximately fifteen minutes.
it will only get worse.
References:
•If greater than 50% of the stem’s circumference is severely compressed, it is probBurns, Russell M., and Barbara H. Honkala, tech. coords.
1990. Silvics of North America: 2. Hardwoods.
ably best and safest to remove the tree and
Agriculture Handbook 654. U.S. Department of
start over.
Agriculture, Forest Service, Washington, DC.
Treatments for affected trees are uncertain. If
SERs (stem encircling roots) can be removed
before compression begins, that’s an excellent and effective treatment. If the SERs
have become SGRs and if, during the course
of removing SGRs, the stem is wounded, the
long-term potential for recovery is uncertain.
The study of stem girdling roots is a relatively
young science and long-term data on treatment options and efficacy are not there. If
50% or more of the tree’s trunk is severely
compressed by the SGRs, and if the symptoms
included dieback and severe stunt, the tree is
probably beyond salvation. If that same tree is
ten feet from a house or utility line, then the
risk of leaving the tree is unacceptable. Buy a
new tree. Remove the excess soil over the root
system. Plant it with the trunk fully exposed.
Mulch the roots, not the trunk. These steps
will put your new tree well on the way to a
long, healthy life.
Johnson, Gary and Dennis Fallon. 2007. Stem Girdling
Roots: The Underground Epidemic Killing
our Trees. www.forestry.umn.edu/extension/.
Under “Hot Topics” on the home page, access
“Stem girdling roots.”
Johnson, Gary and Richard Hauer. 2000. A Practitioner’s
Guide to Stem Girdling Root of Trees: Impacts
on Trees, Symptomology, and Prevention.
www.forestry.umn.edu/extension/. Access the
“Urban Forestry” folder, then “Stem Girdling
Roots” under “Tree Care.”
Moll, Gary. 1989. The state of our urban forest. American
Forests. 95(11/12):61–64.
Summary of Storm Damage to Urban Trees in Minnesota:
1995–2005. www.forestry.umn.edu/extension/.
Access the “Research” folder, then “Storm
Damage 1995–2005.”
USDA Forest Service. 1998. Urban Forest Health:
Identifying Issues and Needs within the
Northeastern Area.
Gary Johnson is a professor in the Department of Forest
Resources at the University of Minnesota.
Photo by Dave Hanson
10 Arnoldia 66/2